-
Notifications
You must be signed in to change notification settings - Fork 3.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Relay] More type alpha equality test coverage #1823
Conversation
t2 = relay.TensorType((1, 2, 3), "float32") | ||
t3 = relay.TensorType((1, 2, 3, 4), "float32") | ||
|
||
tr1 = relay.TypeRelation(None, tvm.convert([t1, t2]), 1, None) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I would really appreciate pointers on how to specify a type relation function or attributes node from within Python (can't find any examples), if possible
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Excellent, thank you. I will make the relation examples less trivial then.
415e047
to
e39e3d9
Compare
Given the operator push, I'll leave expr alpha equality tests for another PR |
Alpha equality is deeply relied on by much of Relay, yet we have very little test coverage for it (mostly lost across different rewrites). I am going to add tests for more cases and in the process try to fix any missing functionality in alpha_eq itself. Please review my tests, as well as changes to alpha_eq, or suggest ways to improve or better organize the tests.
Still to do: